1
Running head: European nightjar home-range and habitat use
2
Home-range size and habitat use of European Nightjars (Caprimulgus
europaeus) nesting in a complex plantation-forest landscape
3
4
5
KATRINA SHARPS, 1* IAN HENDERSON,2 GREG CONWAY,2 NEAL ARMOUR-
6
CHELU3 & PAUL M. DOLMAN1
7
1
8
2
9
3
School of Environmental Sciences, University of East Anglia, Norwich, Norfolk, NR4-7TJ, UK.
British Trust for Ornithology, The Nunnery, Thetford, Norfolk, IP24- 2PU, UK.
Forestry Commission, East England, Santon Downham, Brandon, Suffolk, IP27- 0TJ, UK.
10
11
*Corresponding author:
12
Email: katrinasharps1@gmail.com
13
1
14
In Europe, consequences of commercial plantation management for bird species of
15
conservation concern are poorly understood, in contrast to the strong evidence base for
16
farmland birds. The European Nightjar (Caprimulgus europaeus) is of conservation concern
17
across Europe due to population depletion through habitat loss. Pine plantation-forest is now a
18
key nightjar nesting habitat, particularly in North-Western Europe and increased
19
understanding of foraging habitat selection is required. We radiotracked 31 nightjars in an
20
extensive (185km2), complex conifer plantation landscape in 2009 and 2010. Home-range 95%
21
kernels for females, paired males and unpaired males were an order of magnitude larger than
22
song territories of paired males, emphasising the importance of habitats beyond the song
23
territory. Nightjars travelled a mean maximum distance of 747m from the territory centre per
24
night. Home range placement relative to study landscape composition was examined by
25
compositional analysis. Pre-closure canopy forest (aged 5-10 years) was selected at all scales
26
(MCP, 95% and 50% kernel), with newly planted forest (aged 0-4 years) also selected within
27
50% kernels. For telemetry fixes relative to habitat composition within 2 km of their territory
28
centre, individuals again selected pre-closure and newly-planted forest, and also grazed grass
29
heath. Open ungrazed habitat was not selected, with implications for open habitat planning for
30
biodiversity conservation within public owned forests. Despite nightjar habitat selection, moth
31
biomass was greater in older forest stands, suggesting that foraging site selection reflects ease
32
of prey capture rather than solely prey abundance. Within large plantation-forest landscapes,
33
a variety of growth stages is important for this species of conservation concern and we
34
recommend grazing of open habitats within and adjacent to forest to additionally benefit the
35
European Nightjar.
36
37
Keywords: habitat selection, foraging, radiotracking, moths
38
2
39
Commercial forest management has important implications but also potential for
40
temperate biodiversity (Peterken 1996; Wallace 2011). Forest age structure, structural
41
complexity and the availability of open habitats determine the species and trait composition of
42
invertebrate (Koivula & Niemelä 2003; Pedley et al. 2013) and plant assemblages (Eycott et al.
43
2006). But the implications of European forest management for bird assemblages are not well
44
understood, in marked contrast to the robust evidence base regarding effects of agricultural
45
management and mitigation measures on farmland birds (e.g. Donald et al. 2001). A body of
46
research has examined the effects of area and landscape context on forest bird assemblages
47
(reviewed by Dolman et al. 2007; Dolman 2012) and avian responses to coppice management
48
in semi-natural deciduous woodlands (e.g.Fuller et al. 2007; Holt et al. 2010). However, within
49
managed coniferous plantation forests, avian responses to landscape composition and stand
50
level management have received scant attention, (though see Donald et al. 1998; Fuller et al.
51
2007), despite their importance to numerous bird species of conservation concern, including
52
species otherwise associated with shrubland, heathland or farmland habitats.
53
During the 20th Century, the European Nightjar suffered a widespread population and
54
range reduction across Europe, primarily due to habitat loss, and was subsequently
55
categorised as a Species of European Conservation Concern (SPEC 2) (Burfield & Van Bommel
56
2004) and protected under Annex 1 of the EC Birds Directive (EC 1979). European Nightjars
57
breed in dwarf-shrub heathland and in pine plantation with sparse tree cover, particularly in
58
North-Western Europe (Tucker et al. 1994); however the extent to which populations breeding
59
in conifer forests also depend on other foraging habitats in the wider landscape is not clear.
60
Improved understanding of the foraging behaviour of forest-nesting nightjars, including home-
61
range extent, the relative importance of forest, heathland and other open habitats, and factors
62
driving habitat selection, will provide an evidence basis for conservation management.
63
Although nightjar song territories (including the nest-site) in plantation-forests have
64
been characterised (Ravenscroft 1989, Scott et al. 1998), home-range extent and habitat use
3
65
beyond this are less understood. Nightjars may forage short distances from the nest,
66
particularly when they have eggs or young (Schlegel 1967, Cross et al. 2005), but have also
67
been recorded foraging an average distance of 3.1km per night (Alexander & Cresswell 1990).
68
Animal home range size may vary between sexes, due to differences in behaviour (e.g. Gray et
69
al. 2009). Unpaired male birds can behave differently from paired males, for example
70
defending multiple territories (Amrhein et al. 2007). The female nightjar’s home-range has not
71
been quantified, results presented for males are based on few individuals (e.g. n = 3, Sierro et
72
al. 2001, n = 4, Spray 2007) and no study has distinguished between paired and unpaired birds.
73
Radiotracking studies of nightjars nesting in plantation forest have produced
74
contrasting results for habitat selection. Birds left mature coniferous plantations to feed in
75
deciduous woodland in Dorset, UK (Alexander & Cresswell 1990) and selected open oak
76
scrubland in preference to vineyards or dense pine forest in the Swiss Alps (Sierro et al. 2001).
77
In contrast, nightjars nesting in mixed age pine plantation foraged in young forest and
78
heathland in East Anglia (Bowden & Green 1991) and in South Wales, UK (Cross et al. 2005),
79
where they also utilised rough pasture. This suggests that nightjar foraging behaviour may vary
80
depending on the availability of habitats within the landscape and on the age structure of the
81
plantation-forest itself. However, previous nightjar radiotracking studies were limited by
82
sample size (n = 11 birds, Alexander & Cresswell, 1990, n = 3 Sierro et al. 2001) or difficult
83
terrain (Cross et al. 2005).
84
Nightjar foraging habitat selection may be based on prey abundance but also the ease
85
of capturing prey (Bowden & Green 1991, Sierro et al. 2001). As moths are a key nightjar prey
86
(Collinge 1920, Schlegel 1967), examination of moth biomass within different habitats,
87
particularly in plantations of different ages and heathland, would enhance current knowledge
88
of nightjar habitat requirements. In the UK, where the Open Habitats Policy aims to recreate
89
open areas within existing forest plantation to benefit species including the nightjar (FE 2013),
90
a clearer understanding of nightjar foraging habitat is crucial.
4
91
In order to increase the knowledge base on the importance of management in
92
coniferous plantations for bird species of conservation concern, we radiotracked male and
93
female nightjars in a complex plantation-forest, comprising a mosaic of tree growth stages and
94
patches of heathland in eastern England, UK. We examined: home-range extent of female,
95
paired males and unpaired males; distances travelled; nightjar habitat use within the
96
landscape and used moth trapping to determine the relationship between habitat selection
97
and available prey biomass.
98
Methods
99
Study site
100
The study was conducted in Thetford Forest (0˚40’E, 52˚27’N), the largest lowland
101
commercial forest in the UK, covering 185km2 of Breckland, East England (Fig.1). This region is
102
characterised by a semi-continental climate and sandy soils and supports many species
103
associated with heathland and ruderal land-uses (Dolman & Sutherland 1992). The forest
104
constitutes part of the Breckland Forest Special Protection Area (SPA), designated under the EC
105
Birds Directive (EC 1979) in 2006 for its internationally important breeding populations of
106
nightjar and woodlark (Lullula arborea). Thetford Forest held 349 males (c. 10% of the UK
107
nightjar population) in 2004 (Conway et al. 2007), representing a decrease from the 1998 total
108
of 420 males (Evans 2002). Numbers have subsequently continued to decline, with 240 males
109
counted in 2010 (Conway & Henderson 2010).
110
Thetford Forest is divided into discrete geographical blocks surrounded by
111
predominantly agricultural land and heathland. All designated heathland is grazed by
112
livestock. Additionally, restoration of seven previously forested areas (totalling 300ha) to
113
grazed heathland began in 2000. Corsican (Pinus nigra) and Scots pine (P. sylvestris) comprise
114
85% of the planted area of the forest (Forestry Commission GIS database). Management by
115
clear-felling (at economic maturity, currently 60-80 years) and replanting of even-aged patches
116
of trees creates a mosaic of growth stages (Eycott et al. 2006). Such contiguous areas of trees
5
117
planted in a single year are hereafter referred to as ‘stands’ (mean area 9.0 ha ±8.6 sd)
118
(Dolman & Morrison 2012). Forest growth stages, following Hemami et al. (2004) (Supporting
119
Information, Table S1), were classified as: restocked (0-4 years since planting); pre-thicket (5-
120
10 years); thicket (11-20 years); pole (21-44 years); and mature conifer (≥45).
121
Radiotracking
122
Nightjars were captured in three distinct areas (Fig. 1), containing similar
123
configurations of habitat, chosen to maximise the number of nightjars captured and based on
124
areas of suitable nesting habitat and accessibility. Birds were located through initial surveys
125
conducted one hour before and after sunset during late May 2009 and 2010. Where nightjars
126
were present, capture was attempted using playback lures (consisting of contact and courtship
127
calls, wing clapping display and male churring) placed at the mid-point of a mist-net. Radio
128
tags, (Biotrack Pip Ag-392) with an above-ground detection range of 0.5-1.2km, were attached
129
to the central tail feather following Bowden and Green (1991). Tags weighed 1.5g, within the
130
recommended threshold of 2% (Kenward 2001) of adult body-weight (mean = 74.25g ±7.14sd,
131
range = 65-95g, n = 33). All tagged birds were ringed, therefore birds were individually
132
recognisable. As females responded less than males to the lure, further attempts were made
133
to catch females in the vicinity of active nests.
134
Nightjars roost during daylight (Cramp 1985) therefore to investigate foraging habitat
135
use, tagged nightjars were tracked from sunset until sunrise between 25th May and 28th
136
August 2009 and 2010 using a Biotrack Sika receiver and 3-element Yagi antenna. Between one
137
and three independent tracking teams operated each night, each comprising two people. Each
138
team tracked one tagged bird per night, with information on the location of other tagged birds
139
in the search area recorded where possible. Birds were tracked using the burst sampling
140
method, following Barg et al. (2005) with bursts lasting approximately eight hours; from just
141
before individuals became active at sunset (c. 21.00) until sunrise (c.05.00) with the last fix
142
taken after the bird was roosting. Fixes were taken every ten minutes from sunset to dusk and
6
143
from dawn to sunrise, and reduced to every 30 minutes between dusk and dawn, when the
144
birds were generally stationary (with the exception of females provisioning chicks throughout
145
the night, which were sampled at a fixed rate of every ten minutes). The interval of ten
146
minutes was sufficient for a nightjar to cross its home-range, ensuring independence of each
147
fix (following Kenward 2001).
148
For each bird, fixes were triangulated sequentially by one person, using compass
149
bearings from three consecutive locations (recorded using hand-held GPS); from an open
150
position whenever possible to reduce signal reflectance (Kenward 2001). The average interval
151
between consecutive bearings was four minutes (±2sd). For each bearing, the time, signal
152
strength and additional clues to the bird’s location were noted (for example, if the bird was
153
sighted or heard). Fluctuations and changes in signal volume indicated that the bird was
154
moving. The bird’s activity was classed as stationary, churring, active or interactive (flying in a
155
group, wing clapping or pursuing another bird) (Supporting Information, (2)). Fix locations
156
were subsequently determined from bearings using LOCATE III (Pacer Co., Canada). Tracking
157
error was measured by comparing locations estimated from triangulation (using either two or
158
three bearings) to the known location of stationary tags (n = 32, measured using handheld GPS
159
with an error < 10m), held by a colleague at ground level in a variety of forest growth stages
160
(restocked, pre-thicket, thicket, mature; n = 8 per habitat).
161
Breeding status
162
Male status (paired or unpaired) was judged using multiple criteria (Supporting
163
Information, Table S4). Although behavioural observations are commonly used to identify the
164
breeding status of male birds (e.g. Van Horn et al. 1995, Guillemain et al. 2003), we
165
acknowledge that some males may have been misclassified or changed status during the
166
season.
167
168
Nightjar ranging and habitat use
7
169
Both male and female nightjars spend a large proportion of time stationary near the
170
nest (Cramp 1985). As home-ranges based on all tracking fixes would be defined by the nest-
171
site rather than foraging locations, only active fixes were used in the analysis. Although
172
criticised for failing to describe space utilisation within the range (Worton 1987), Minimum
173
Convex Polygons (MCPs) are considered comparable among studies (Kenward 2001). Male
174
song territory MCPs were created using fixes for churring locations. Home-range MCPs were
175
also calculated (n = 29, mean fixes = 37, ±16sd, range = 17-72 fixes), with three individuals
176
excluded due to an insufficient number of fixes. The Kernel Density Estimator (Worton 1989) is
177
now accepted as a more biologically meaningful method for home-range analysis (Seaman &
178
Powell, 1996) and was calculated for females, paired and unpaired males using fixed kernels
179
based on a constant proportion of href for all individuals (Bertrand et al. 1996, Kie et al.
180
2002)(see Supporting Information (4)). Females were not split by nesting status due to
181
insufficient sample size. Home-range kernels at 95% (outer range) and 50% (core range) were
182
calculated for all birds with at least 27 active positional fixes (n = 19, mean fixes = 44.6,
183
±15.6sd, range = 27-72) following Seaman et al. (1999). MCPs and kernels were produced in
184
ArcGIS 9.2 using Hawth’s Tools and Home-Range Tool (Rodgers & Kie 2011).
185
Home-ranges analysed in this study were restricted to stable periods for individuals,
186
with consistent breeding status. For four male birds that changed from paired to unpaired
187
breeding status during the season, only one song territory and home-range (that with the most
188
fixes) was analysed. Multiple nesting attempts within the original song territory (n = 3, mean
189
distance from previous nest = 88m, ±76sd) were included in one home-range. One female
190
moved territory centre by 2161m, after fledging a first brood; two MPCs were created for this
191
individual but insufficient fixes were available to calculate two kernels, therefore one large
192
kernel was created. Birds excluded from the kernel analysis included two males that changed
193
breeding status, resulting in insufficient fixes per status class.
8
194
To investigate nightjar foraging distances, the distance between every active fix and
195
the territory centre (roost or nest-site) was calculated using Hawth’s Tools with ArcGIS 9.2.
196
Periods of absence (no signal) during tracking were also recorded. The longest distance
197
travelled from the roost/nest per night for each bird was calculated (excluding nights where
198
the bird had no signal for >10 minutes). Due to extreme values, both mean and median
199
distances are reported.
200
201
202
Moth trapping
In order to compare biomass of a commonly exploited prey (moths) (Collinge 1920,
203
Schlegel 1967) between habitat classes, heath-type 15W actinic moth traps (Anglian
204
Lepidopterist Supplies) were placed in five habitats available to foraging nightjars: grass-heath;
205
ungrazed grassland; restocked; pre-thicket and old trees (including pole and mature trees,
206
range = 21-80 years) on five nights each week (where possible) between the beginning of June
207
to the end of August in 2009 and 2010 (n = 423 trap-nights).
208
The attraction radius for actinic traps is low with few moths recaptured when released
209
more than 40m away (Truxa & Feidler 2012); therefore to sample moths from the target
210
habitat only, traps were placed ≥ 50m from the stand edge. Traps were positioned at sunset
211
and emptied at sunrise, with moths inside the trap or on the box exterior recorded.
212
Temperature was recorded using data loggers (Lascar electronics, El-USB-1). In 2009, three
213
loggers (one each in un-grazed grassland, restock and pre-thicket forest) were placed in the
214
central forest block. In 2010, 20 loggers were deployed in three forest blocks, in five habitats,
215
(grass-heath, un-grazed semi-natural grassland, restock, pre-thicket and mature forest; n = 4
216
per habitat).
217
As moth abundance in Thetford Forest peaks at dusk (Bowden & Green 1991), the
218
daily time for the end of evening twilight (the final stage of dusk) was obtained from
219
www.timeanddate.com (Norwich, UK) and the mean dusk temperature from all data loggers
9
220
was used for the nightly temperature (Supporting Information, (4)). Dry weights of moths
221
collected from each family group were used to determine moth biomass (Supporting
222
Information, (4)).
223
224
Statistical analysis
225
Song territory size (square root transformed) was compared between paired and
226
unpaired males using a General Linear Model (GLM), with male breeding status as a fixed
227
effect, including ‘number of fixes’ as a covariate. Home-range size (square root transformed)
228
was compared among females, paired males and unpaired males using a GLM, again
229
controlling for number of fixes as a covariate. Differences among classes were examined using
230
pair-wise comparisons of estimated marginal means. GLMs were conducted in PASW Statistics
231
v. 17.0.3.
232
Compositional analysis (Aebischer et al. 1993) was used to investigate selection of
233
foraging habitat, initially considering all birds rather than splitting by sex or status, due to
234
limited sample sizes. Habitats within the study area were classified using GIS (Supporting
235
information, Tables S1 &2). Those with < 2% availability within the study area were excluded
236
(scrub; improved grassland; inaccessible (tourist complex); urban and water), retaining for
237
analysis: grazed grass-heath (including heathland within and adjoining the forest; referred to
238
as ‘grass-heath’ hereafter); ungrazed semi-natural grassland; restocked stands; pre-thicket;
239
thicket; pole stage; mature conifer; arable and mature broadleaf (referred to as ‘broadleaf’
240
hereafter).
241
The female nesting habitat, and for male birds, the primary habitat ( > 80%) within the
242
song territory, were extracted using GIS. Nightjar foraging habitat selection was investigated at
243
two levels of spatial scale following Johnson (1980); considering home-range placement (MCP,
244
95%, 50% kernels) relative to habitat available within the study area and individual fix
245
locations, relative to the habitat available to that individual.
10
246
For the first level, as tagged birds were aggregated in different blocks of the forest,
247
three discrete study area MCPs were created based on pooled fixes across individuals from
248
both years (Fig. 1). To encompass habitat that was potentially available to the birds, each study
249
area MCP was buffered by 1.2km, based on distances travelled by nightjars in this and previous
250
studies (Alexander & Cresswell 1990, Bowden & Green 1991), resulting in a mean distance
251
from the individual territory centre to the buffered edge of the study area of 1.98km (±0.56
252
sd).
253
For the second stage of analysis, habitat available to each individual was considered as
254
that within a fixed radius (2km) of the territory centre. This was considered preferable to using
255
individual MCPs or kernels as: 1) home-range size varied more than tenfold among individuals;
256
2) h ref kernel estimators produced multi-modal home-ranges that delineated used habitat and
257
excluded much of the traversed habitat not used for foraging, thus precluding examination of
258
used vs. available at the scale of the fix. The distance chosen was consistent with the study
259
area buffers and encompassed 100% of fixes for 2009 and 98% for 2010. The percentages of
260
each habitat class available to birds and of fixes in each habitat were calculated using ArcGIS
261
9.2, excluding inaccessible areas.
262
Hierarchical habitat selection was examined using ‘Compos Analysis Version 6.2 Plus’
263
(Smith Ecology 2004). A habitat preference index was created based on the sum of the
264
difference in log ratios between habitats produced from the compositional analysis (following
265
Holt et al. 2010). Habitat use values of zero were substituted with a number an order of
266
magnitude smaller than the values for available and used habitat (Aebischer et al. 1993) and
267
1000 iterations were chosen for data randomisation. Individuals were weighted according to
268
the number of fixes obtained (Aebischer et al. 1993).
269
As birds were not split by status for the analysis of home-range kernel placement
270
within the study area, two males excluded from the kernel home-range area analysis (due to
271
changed breeding status) were included. When considering individual fixes within a 2km
11
272
buffer, four birds were excluded as they had no access to ‘grass-heath’ and three further birds
273
were excluded as there were insufficient fixes to draw meaningful results (fewer than five). As
274
it was difficult to differentiate between male foraging activity and territorial behaviour, the
275
compositional analysis containing all birds may be biased towards habitats within the song
276
territory. Female birds were considered more likely to be foraging when active, due to
277
constraints of incubation and chick rearing, relative to territorial males; therefore analyses at
278
scales of MCP placement and individual fixes were repeated using female data only. Analysis of
279
female kernel placement relative to available habitat was not possible as the number of
280
habitats (n = 9) was greater than the number of female kernels (n = 7), violating compositional
281
analysis rules (Aebischer et al. 1993).
282
Moth biomass (square root transformed) was compared among habitats using a
283
General Linear Mixed Model (GLMM) with Gaussian error, including dusk temperature and
284
temperature2 as fixed effects, an interaction between habitat type and temperature, and
285
forest stand as a random effect. Model selection was carried out by examining the change in
286
the Akaike Information Criterion corrected for small sample size (AICc) on variable removal,
287
following Burnham and Anderson (2002). To investigate the difference in moth biomass among
288
habitat types, we examined whether 95% confidence intervals of the habitat coefficient
289
estimates overlapped zero, sequentially changing the reference habitat to undertake pairwise
290
comparisons (following Boughey et al. 2011). Mixed models were conducted using package
291
lme4 in R (R Development Core Team 2012).
292
293
294
Results
Thirty-six birds were radiotagged; 20 in 2009 (13 male and seven female) and 16 in
295
2010 (12 male and four female). In 2009, two birds shed their tags and three were lost, leaving
296
a total of 31 (21 male and ten female) for analysis. No individuals were tracked in both
12
297
seasons. The total recording effort (duration of nightly fieldwork) was 1518 hours. In trials with
298
stationary tags, mean tracking error was 26m (± 20sd, range = 2-99m) and did not differ with
299
bearing number (F1, 56 = 0.317, p = 0.576) or forest growth stage (F3,56 = 1.237, p = 0.305). This
300
error may vary depending on the distance and activity of the bird (Supporting Information (2)).
301
Of the tagged males, 11 were initially paired and ten were considered to be unpaired, with
302
four of the paired males becoming unpaired during the season. Two of the eight unpaired
303
males included in the analysis had extensive song territories, with one bird recorded churring
304
3.6km away from the territory centre (on one occasion) and a second bird increasing the size
305
of its song territory (fourfold) towards the latter half of the breeding season. Two females
306
moved territory, after the fledging (2161m) or failure (819m), of the first nest , the latter
307
before regular radiotracking began.
308
309
310
Ranging of nightjars
Song territory size did not differ between paired males and those considered unpaired
311
(Fig. 2. a) and was positively related to the number of fixes (F1, 16 = 5.2, p = 0.03). Home-range
312
area for MCPs, 95% and 50% kernels increased with the number of fixes (F1, 25 = 8.9, p < 0.01,
313
F1, 15 = 12.5, p < 0.01, F1, 15= 14.6, p < 0.01 respectively). Controlling for fix number, home-range
314
MCP and 95% kernels for males were an order of magnitude larger than the song territories of
315
paired males (Fig. 2. b & c). Females had home-ranges twice the size of paired males in terms
316
of MCP and over three times the size for 95% and 50% kernels (Fig. 2) Home-range size (MCP
317
and kernels) for unpaired males was intermediate and did not differ from female or paired
318
male ranges (Fig. 2. b, c & d). While the 50% kernel for all birds included the nest/song
319
territory, females had multi-modal 50% kernels and paired males tended to have one central
320
50% range within the song territory (Supporting information, Fig. S1).
321
322
Active fixes (n = 1099 fixes) were a mean of 507m (± 482sd, range = 3-3225m), and
median of 344m from the territory centre. Of the total number of attempted active fixes (n =
13
323
1350), 10% resulted in the bird recorded missing (i.e. no signal) for > 10 minutes. Birds were
324
also recorded missing for shorter periods ( < 10% of attempted fixes), considered unlikely to
325
represent foraging to remote locations. Of all missing signals, most (70%) were from just eight
326
of 27 birds. The most and longest absences where the location of the bird remained unknown,
327
but may have been external to the forest, were for unpaired males and two females that
328
ranged widely. the mean of the nightly maximum distance that birds were detected from the
329
territory centre was 747m (± 513sd, range = 72-2601m, n=111 nights) and the median nightly
330
maximum was 575m. Of the total bird tracking nights (n = 208), 34% were excluded from the
331
calculation of maximum distance travelled due to missing signals (25% with one or more
332
absences > 15mins, 9% with one or more absences of 11-15mins); 13% were excluded due to
333
other factors (e.g. the tracked bird was not the primary bird for the night).
334
335
336
Habitat selection
Nest and song territories were recorded in pre-thicket (n = 15 birds), restocked (n = 9)
337
and thicket (n = 5) habitats. For home-range MCPs, relative to availability in the study area,
338
pre-thicket plantation had the highest habitat preference index and was used significantly
339
more than grass-heath, ungrazed grassland, arable, broadleaf and restocked stands. No
340
significant difference was found among pre-thicket, mature, pole and thicket forest stages (Fig.
341
3a). For 95% kernels, pre-thicket was again ranked the most selected, with no significant
342
difference among the forest habitats (except for mature), however pre-thicket was used
343
significantly more than grass-heath, ungrazed grassland, arable and broadleaf (Fig. 3b). Within
344
the 50% core range, pre-thicket was the most selected, followed by restocked stands, with
345
ungrazed grassland, arable and broadleaf being the least selected (Fig. 3c). For habitat use at
346
individual fixes within the 2km buffer, pre-thicket, restocked and grass-heath were selected
347
(no significant difference among these), while there was no difference between ungrazed
14
348
grassland, pole, thicket and mature, with arable and broadleaf ranked as the least selected
349
(Fig. 3d).
350
Compositional analysis restricted solely to females showed a similar pattern of habitat
351
selection (Fig. 3a &d), with pre-thicket the most strongly selected at all stages of analysis.
352
However, low sample size resulted in fewer significant differences among habitats.
353
354
355
Moth biomass among habitats
Moth biomass in old trees was more than double that of young forest (both pre-
356
thicket and restock) and ungrazed open habitats, while biomass in grass-heath sites did not
357
differ from that in either old trees, or young forest (Fig. 4). More moth biomass was captured
358
on warmer nights (Δ AICi for model lacking temperature and temperature2 = 246). Old stands
359
had greater moth biomass than other habitats on nights with low temperatures (Δ AICi for
360
model lacking the interaction between habitat and temperature = 8).
361
362
363
Discussion
The home-range of the European Nightjar in mixed plantation was an order of
364
magnitude larger than the song territories of paired males. However, the mean recorded
365
maximum foraging distances per night were shorter than those recorded in other UK nightjar
366
populations. Results show that a complex forest landscape can provide the nightjar with
367
habitat for both nesting and foraging and the outcomes of this study can be used to inform
368
habitat management.
369
370
371
Nightjar ranging behaviour
The home-ranges in this study were based on active fixes, with the majority assumed
372
to represent foraging. However, social contact may have also influenced ranging behaviour.
373
These activities may not be mutually exclusive however, with territorial males recorded
15
374
occupying the song territory all night, only making short flights between churring posts, thus
375
birds were presumed to be foraging while defending the territory. While song territory size did
376
not differ between paired and unpaired males, two (of eight) unpaired males were recorded
377
churring outside of the song territory towards the end of the season. Thus, although unpaired
378
nightjars do hold a territory, they may leave or extend their song territory as the season
379
progresses, as do nightingales (Luscinia megarhynchos) (Amrhein et al. 2007).
380
The larger size of the home-ranges confirmed that nightjars are leaving the song
381
territory, it is assumed, to forage. Female 95% home-range kernels were larger than those of
382
paired males, with wide-ranging active fixes, which may represent travel to optimal foraging
383
grounds. One female had a large kernel primarily as a result of moving to a new territory.
384
Cresswell and Alexander (1990) document a case of a female nightjar swapping partners
385
between nesting attempts, therefore mate prospecting may account for some of the outer
386
home-range points of other females. While female 50% kernels were large due to their multi-
387
modal structure, for paired males, the core (50%) area was centred on the song territory. It
388
may be that while a priority for the male is to guard the territory, for females it is to find
389
optimal foraging sites. Unpaired males defended a territory but ranged further than paired
390
males, perhaps to increase the chance of locating a female. All birds, irrespective of sex and
391
status, were regularly active in the song/nest territory, suggesting that nightjars in plantation-
392
forest may choose a territory based on its suitability for both nesting and foraging.
393
While Alexander and Cresswell (1990) recorded a mean maximum distance per night
394
of 3.1 ±1.2 km, the mean maximum distance for nightjars in Thetford Forest was 747m±513sd.
395
Due to challenges of relocating birds within the forest and surrounding landscape when their
396
position was unknown, birds were recorded as out of range for > 10 minutes at 10% of
397
attempted active fixes, and on 34% of tracking nights. Such points may have been external to
398
the forest but mostly related to unpaired males and two females (one after moving nest
399
territory, another after the failure of the second nest); these birds may not have been solely
16
400
foraging. While these absences must be taken into consideration, nightjars nesting in mixed
401
pine-plantation were not found to leave coniferous forest on the majority of foraging trips, in
402
contrast to patterns found in other populations (Alexander & Cresswell 1990, Sierro et al.
403
2001).
404
405
406
Nightjar habitat selection
For the placement of MCPs within the study area, there was no significant difference
407
in habitat selection among pre-thicket, thicket, mature and pole stage forest; however, this
408
may be because MCPs included areas alongside or between key habitats. Similarly, when 95%
409
kernels were considered relative to the study area, there was no significant difference among
410
the majority of forest growth stages. As the forest contains a mosaic of growth stages and
411
kernels represent the likelihood of locating the bird 95% of the time (birds were recorded
412
active at least once in all forest growth stages), it may have been difficult to distinguish
413
between forest habitats at this level of analysis.
414
Pre-thicket stage forest was found to be the most strongly selected habitat for all
415
stages of compositional analysis, with restock also selected within 50% kernels, relative to the
416
study area. The key foraging habitats in this study are similar to preferred nightjar song
417
territory habitats (young forest aged 0-10 years) (e.g. Ravenscroft 1989). The majority of
418
territories (nest and song) in the current study were based in restocked and particularly pre-
419
thicket habitats and it was difficult to separate foraging and male territorial activity. It is
420
therefore possible that the compositional analysis including all birds may be biased by song
421
territory location. However, when the females were off the nest and more likely to be foraging
422
(compared to male birds), they showed similar habitat selection.
423
Nightjars selected grazed grass-heath (relative to availability) when available within a
424
2km radius. Livestock dung in grazed habitats may provide beetle prey for nightjars (Schlegel
425
1967). In contrast, ungrazed grassland was one of the least selected habitats. Similarly,
17
426
Alexander and Cresswell (1990) recorded foraging nightjars avoiding ungrazed Calluna
427
dominated heathlands. Nightjars have been recorded ‘hawking’ for prey (Cramp 1985), which
428
involves swooping for flying insects from (or close to) the ground. This may be more difficult in
429
taller, denser ungrazed grasslands.
430
While nightjars have been found to regularly leave territories within conifer forest to
431
feed in atypical habitats (Alexander & Cresswell 1990, Sierro et al. 2001), these forests were of
432
uniform age and densely planted, in contrast to the mixed-age structure of Thetford Forest.
433
Broadleaf woodland was ranked as one of the least selected habitats in the current study, but
434
was the preferred nightjar foraging habitat in Dorset, UK (Alexander & Cresswell 1990). This
435
may be because broadleaf patches in Thetford Forest primarily comprised dense mature
436
sycamore (Acer pseudoplatanus) or beech (Fagus sylvatica) rather than mixed semi-natural
437
deciduous woodland. Overall, results suggest that habitat structure may influence the extent
438
to which nightjars forage in the wider forest landscape.
439
440
Moth biomass in habitats
441
Nightjars in Thetford Forest were not choosing habitats based solely on moth
442
abundance as stands containing old trees (pole and mature) had a greater moth biomass than
443
younger stands. Older forest growth stages were found to have relatively greater moth
444
biomass even on cold nights, presumably due to denser canopy cover keeping temperatures
445
higher than in other growth stages. This supports the inference that nightjars select habitat
446
based on the ease of prey capture rather than abundance, with older forest growth stages
447
avoided due to the difficulty of finding and capturing prey amongst dense branches (Bowden &
448
Green 1991, Sierro et al. 2001).
449
Habitats may also be selected due to the abundance of other prey types such as
450
beetles, another key component of the nightjar diet (Cramp 1985; Sharps 2013). Beetles may
451
be an important prey both in restocked and pre-thicket stands due to the abundance of
18
452
saproxylic beetles (e.g. Cerambycidae) emerging from cut stumps (Hedgren 2007), and in
453
grazed grass-heath, which supports dung-feeding beetles (Aphodiinae) (Menéndez and
454
Gutiérrez 1996), although abundant deer populations (Wäber & Dolman 2013) also support
455
dung-feeding beetles in forest habitats (Stewart 2001).
456
457
458
Conservation implications and recommendations
Understanding of European Nightjar habitat use in plantation-forest provided by this
459
study has widespread application, particularly in North-Western Europe. While Gribble (1983)
460
and Morris et al. (1994) showed the importance of pine-plantations to nesting nightjars, this
461
study demonstrates that mixed age plantation-forest can provide both nesting and foraging
462
habitat for this species. Nightjars were found to be foraging in forest habitats exterior to the
463
song territory. Forest-nesting nightjars would therefore benefit most from a large forest
464
comprising a mosaic of growth stages, for which rotational clear-felling and replanting of
465
large, even-aged stands (rather than group selection) is ideal. The configuration of growth
466
stages within the forest mosaic may also be important for the nightjar; stands of young forest
467
clustered together would provide female nightjars with multiple foraging opportunities in close
468
proximity.
469
The results of this study suggest that the grazing of open habitats is important for
470
foraging nightjars, while creation or retention of ungrazed, unplanted patches within the forest
471
may not be beneficial. The introduction of further patches of grazed heathland reversion
472
within and adjacent to the forest would provide additional foraging resources for nightjars and
473
benefits for regional biodiversity (Dolman et al. 2012). In terms of breeding nightjars, the use
474
of grazing to maintain the vegetation structure of heathland will be beneficial as long as
475
sufficient nesting habitat is also provided within c. 2km.
476
477
In summary, this study demonstrates that the management of commercial plantationforest has important implications for maintaining populations of this species of conservation
19
478
concern. The creation of a heterogeneous forest mosaic is important for foraging nightjars,
479
reducing the need for birds to leave the forest to reach suitable foraging habitat and allowing
480
both foraging and nesting habitat for this species to be supported by one land
481
owner/manager.
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
20
497
We thank Alastair Feather, Vivien Hartwell, Kirsten Miller, Elwyn Sharps, Laura Wilkinson and
498
Isobel Winney for assistance with fieldwork. Additional thanks to all radiotracking and nest
499
finding volunteers, including BTO, Forestry Commission, RSPB and Wildlife Trust staff. Financial
500
assistance for this study was provided by the Sita Trust, the Forestry Commission and the John
501
and Pamela Salter Charitable Trust.
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
21
529
References
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
Aebischer, N.J., Robertson, P.A. & Kenward, R.E. 1993. Compositional analysis of habitat use
from animal radio-tracking data. Ecology 74: 1313–1325.
Alexander, I. & Cresswell, B. 1990. Foraging by Nightjars Caprimulgus europaeus away from
their nesting areas. Ibis 132: 568–574.
Amrhein, V., Kunc, H.P., Schmidt, R. & Naguib, M. 2007. Temporal patterns of territory
settlement and detectability in mated and unmated Nightingales Luscinia
megarhynchos. Ibis 149: 237–244.
Barg, J.J., Jones, J. & Robertson, R.J. 2005. Describing breeding territories of migratory
passerines: suggestions for sampling, choice of estimator, and delineation of core
areas. J. An. Ecol. 74: 139–149.
Bertrand, M.R., DeNicola, A.J., Beissinger, S.R. & Swihart, R.K. 1996. Effects of parturition on
home ranges and social affiliations of female white-tailed deer. J. Wildl. Manage. 60:
899–909.
Boughey, K.L., Lake, I.R., Haysom, K.A.& Dolman, P.M. 2011. Improving the biodiversity
benefits of hedgerows: how physical characteristics and the proximity of foraging
habitat affect the use of linear features by bats. Biol. Conserv.144: 1790-1798.
Bowden, C. G. R. & Green, R.E. 1991. The ecology of nightjars on pine plantations in Thetford
forest. RSPB Research Department; unpublished report.
Burfield, I. & Van Bommel, F. 2004. Birds in Europe: Population Estimates, Trends and
Conservation Status. Cambridge: BirdLife International.
Burnham, K.P.& Anderson, D.R. 2002. Model Selection and Multimodel Inference: A Practical
Information-Theoretic Approach. New York: Springer.
Collinge, W.E. 1920.The food of the nightjar (Caprimulgus europaeus). J. Min. Agric. 26: 992995.
Conway, G. & Henderson, I. 2010. Breckland Forest nightjar Caprimulgus europaeus survey
2010. BTO Research Report No. 569. British Trust for Ornithology, Thetford.
Conway, G., Wotton, S., Henderson, I., Langston, R., Drewitt, A. & Currie, F. 2007. Status and
distribution of European Nightjars Caprimulgus europaeus in the UK in 2004. Bird
Study 54: 98–111.
Cramp, S. (ed.) 1985. The Birds of the Western Palaertic, Vol. 4. Oxford: Oxford University
Press.
Cresswell, B., & Alexander, I. 1990. A case of mate-swtiching between broods in the Nightjar.
Ring. Migr.11: 73-75.
Cross, T., Lewis, J., Lloyd, J. Morgan, C. & Rees, D. 2005. Science for conservation
management: European Nightjar Caprimulgus europaeus. Breeding success and
foraging behaviour in upland coniferous forests in Mid-Wales. Countryside Council for
Wales; unpublished report.
Dolman, P.M. 2012. Mechanisms and processes underlying landscape structure effects on
bird populations. In: Fuller, R.J. (ed). Birds and Habitat: Relationships in Changing
Landscapes. Cambridge: Cambridge University Press. pp. 93-124.
Dolman, P.M., Hinsley, S.A., Bellamy, P.E., Watts, K. 2007. Woodland birds in patchy
landscapes: the evidence base for strategic networks. Ibis 149 (suppl 2) 146-160.
Dolman, P.M. & Morrison, C. 2012.Temporal change in territory density and habitat
quality for Breckland Forest SSSI woodlark and nightjar populations. School of
Environmental Sciences, University of East Anglia, Norwich.
Dolman, P.M., Panter, C.J., Mossman, H.L. 2012. The Biodiversity Audit Approach challenges
regional priorities and identifies a mismatch in conservation. J. Appl. Ecol. 49: 986–997.
Dolman, P.M. & Sutherland, W.J. 1992. The ecological changes of Breckland grass heaths and
the consequences of management. J. Appl. Ecol. 29: 402–413.
22
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
Donald, P. F.,Fuller, R.J. Evans, R.J. & Gough, S.J. 1998. Effects of forest management and
grazing on breeding bird communities in plantations of broadleaved and coniferous
trees in western England. Biol. Cons. 85: 183-197.
Donald, P. F., Green, R.E. & Heath, M.F. 2001. Agricultural intensification and the collapse of
Europe's farmland bird populations. Proc. Royal Soc. B: Biol. Sci. 268: 25-29..
EC. 1979. Directive on the Conservation of Wild Birds 79/409/EEC.
Evans, M. 2002. GIS-based modelling of woodlark (Lullula arborea) and nightjar (Caprimulgus
europaeus) habitats in a forested landscape. PhD thesis, University of East Anglia,
Norwich, UK.
Eycott, A.E., Watkinson, A.R. & Dolman, P.M. 2006. Ecological patterns of plant diversity in a
plantation forest managed by clearfelling. J. Appl. Ecol. 43: 1160–1171.
FE 2013. A Strategy for Open Habitat Policy Delivery on the Public Forest Estate. Forest
Enterprise, England .
Fuller, R. J., Smith, K.W., Grice, P.V., Currie, F.A. & Quine, C.P. 2007. Habitat change and
woodland birds in Britain: Implications for management and future research. Ibis 149:
261-268.
Gray, T.N., Chamnan, H., Collar, N.J. & Dolman, P.M. 2009. Sex-specific habitat use by a
lekking bustard: conservation implications for the critically endangered Bengal Florican
(Houbaropsis bengalensis) in an intensifying agroecosystem. The Auk 126: 112–122.
Gribble, F. 1983. Nightjars in Britain and Ireland in 1981. Bird Study 30: 165–176.
Guillemain, M., Caldow, R.W.G., Hodder, K.H. & Goss-Custard, J.D. 2003. Increased vigilance
of paired males in sexually dimorphic species: distinguishing between alternative
explanations in wintering Eurasian Wigeon. Behav. Ecol. 14: 724–729.
Hedgren, P.O. 2007. Early arriving saproxylic beetles (Coleoptera) and parasitoids
(Hymenoptera) in low and high stumps of Norway spruce. For. Ecol. Manage. 241:
155–161.
Hemami, M.R., Watkinson, A.R &, Dolman, P.M. 2004. Habitat selection by sympatric muntjac
(Muntiacus reevesi) and roe deer (Capreolus capreolus) in a lowland commercial pine
forest. For. Ecol. Manage. 194: 49–60.
Holt, C.A., Fuller, R.J. & Dolman, P.M. 2010. Experimental evidence that deer browsing
reduces habitat suitability for breeding Common Nightingales Luscinia megarhynchos.
Ibis 152: 335–346.Johnson, D.H. 1980. The comparison of usage and availability
measurements for evaluating resource preference. Ecology 61: 65–71.
Kenward, R. 2001. A Manual for Wildlife Radio Tagging. London: Academic Press.
Kie, J.G., Bowyer, R.T., Nicholson, M.C., Boroski, B.B.& Loft, E.R. 2002. Landscape
heterogeneity at differing scales: effects on spatial distribution of mule deer. Ecology
83: 530–544.
Koivula, M., & Niemelä J. 2003. Gap felling as a forest harvesting method in boreal forests:
Responses of carabid beetles (Coleoptera, Carabidae). Ecography 26: 179-187.
Menéndez, R. & Gutiérrez, D. 1996. Altitudinal effects on habitat selection of dung beetles
(Scarabaeoidea: Aphodiidae) in the northern Iberian peninsula. Ecography 19: 313–
317.
Morris, A., Burges, D. & Fuller, R.J. 1994. The status and distribution of Nightjars Caprimulgus
europaeus in Britain in 1992. Bird Study 41: 181–191.
Pedley, S. M., Franco, A.M.A., Pankhurst, T. & Dolman, P.M. 2013. Physical disturbance
enhances ecological networks for heathland biota: A multiple taxa experiment. Biol.
Cons. 160: 173-182.
Peterken, G. F. 1996. Natural Woodland: Ecology and Conservation in Northern Temperate
Regions. Cambridge: Cambridge University Press.R Development Core Team. 2012. R:
A language and environment for statistical computing. R Foundation for Statistical
Computing, Vienna, Austria.
23
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
Ravenscroft, N.O.M. 1989. The status and habitat of the Nightjar Caprimulgus europaeus in
coastal Suffolk. Bird Study 36: 161-169.
Rodgers, A.R. & Kie, J.G. 2011. HRT: Home Range tools for ArcGIS, A User’s Manual. Centre for
Northern Forest Ecosystem Research, Ontario.
Seaman, D.E., Millspaugh, J.J., Kernohan, B.J., Brundige, G.C., Raedeke, K.J.& Gitzen, R.A.
1999. Effects of sample size on kernel home range estimates. J. Wild. Manage. 63:
739–747.
Seaman, D.E. & Powell, R.A. 1996. An evaluation of the accuracy of kernel density estimators
for home range analysis. Ecology 77: 2075–2085.
Schlegel, R. 1967. Die Ernahrung des Ziegenmelkers Caprimulgus europaeus, seine
wirtschaftliche Bedeutung und seine Siedlungsdichte in einem Oberlausitzer
Kiefernrevier. Beitr. Vogelk., 13: 145-190.
Scott, G.W., Jardine, D.C., Hills, G. & Sweeney, B. 1998. Changes in Nightjar Caprimulgus
europaeus populations in upland forests in Yorkshire. Bird Study 45: 219–225.
Sharps, K 2013. The conservation ecology of the European Nightjar (Caprimulgus europaeus) in
a complex heathland-plantation landscape. PhD thesis, University of East Anglia,
Norwich, UK.
Sierro, A., Arlettaz, R., Naef-Daenzer, B., Strebel, S. & Zbinden, N. 2001. Habitat use and
foraging ecology of the nightjar (Caprimulgus europaeus) in the Swiss Alps: towards a
conservation scheme. Biol. Conserv. 98: 325–331.
Spray, S. 2007. Dumfries and Galloway Nightjar radio tracking project 2006. Stuart Spray
Wildlife Consultancy; unpublished report.
Stewart, A.J.A. 2001. The impact of deer on lowland woodland invertebrates: a review of the
evidence and priorities for future research. Forestry 74: 259–270.
Truxa, C. & Fiedler, K. 2012. Attraction to light - from how far do moths (Lepidoptera) return
to weak artificial sources of light? Europ. J. Entom. 109: 77–84.
Tucker, G.M., Heath, M.F., Tomialojc, L., Grimmett, R.F. & Socha, C.M. 1994. Birds in Europe:
Their Conservation Status. Cambridge: BirdLife International.
Van Horn, M.A., Gentry, R.M. & Faaborg, J. 1995. Patterns of Ovenbird (Seiurus aurocapillus)
pairing success in Missouri forest tracts. The Auk 112: 98–106.
Wäber, K., Dolman, P.M. 2013. Achieving landscape-scale deer management for biodiversity
conservation: the need to consider sources and sinks. J. Wildl. Manage. 77: 726-736.
Wallace, E.W. 2011. Woodlands: Ecology, Management & Conservation. New York: Nova
Science Publishers.
Worton, B.J. 1987. A review of models of home range for animal movement. Ecol. Model. 38:
277–298.
Worton, B.J. 1989. Kernel methods for estimating the utilization distribution in home-range
studies. Ecology 70: 164–168.
24
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
Supporting Information
(1): Habitat classification
Table S1: Classification of forest growth stages
Table S2: Classification of non-forest habitat
(2): Radiotracking methodology
Table S3: Monitoring information for tagged nightjars
(3): Male breeding status
Table S4: Male breeding status criteria
(4): Nightjar home-range kernels
Figure S1. Nightjar home-range kernel density estimates
(5): Moth trapping
25
685
686
Figure 1: Thetford Forest, showing the three study areas; ● = territory centres of radiotagged
687
nightjars; ‘Other habitats’ within the study areas = mature conifer exterior to Thetford Forest;
688
urban; un-grazed semi-natural grassland; improved grassland; broadleaf woodland and water.
689
690
691
692
693
26
694
695
Figure 2. Mean area used by nightjars in Thetford Forest, showing (a) song territory MCP for
696
paired male (n = 11) and unpaired males (n = 8); (b) home-range MCP for female (n = 11),
697
paired male (n = 10), unpaired male (n = 8); (c) 95% kernel and (d) 50% kernel for female (n =
698
7), paired male, (n = 7), unpaired male, (n = 5). Model-derived means are back transformed
699
from square rooted data. Error bars represent 95% CIs . Within each plot, classes sharing a
700
lower case letter do not differ significantly (p>0.05).
701
27
702
703
Figure 3. Index of habitat preference for nightjars in Thetford Forest, (a) MCP home-ranges
704
(All birds, n = 29; Females, n = 11); (b) 95% home-range kernel (n = 21); (c) 50% core kernel (n
705
= 21), each compared with habitat availability across the study area; (d) fixes for each bird
706
compared with 2km buffers around territory centre (All birds, n = 25; Females, n = 10). Habitat
707
types: A = arable, G = grazed grass-heath, UNG = ungrazed semi-natural grassland, RS =
708
restocked forest (0-4 years), PTh = pre-thicket forest (5-10 years), Th = thicket forest (11-20
709
years), P = pole forest (21-44 years), M = mature conifer ( ≥ 45), Br = Broadleaf. Within each
710
plot, classes sharing a super-script do not differ significantly (p > 0.05; determined by
711
compositional analysis) (letters above bars = all birds, letters below the bars = female).
712
28
713
714
715
716
Figure 4. Model-derived mean moth biomass in different habitats. Habitat types: G = grazed
717
grass-heath; UNG = ungrazed semi-natural grassland; RS = restocked forest (0-4 years); PTh =
718
pre-thicket stage forest (5-10 years); Old =pole and mature forest (21-80 years). Means are
719
back transformed from square rooted data. Error bars are ± 95% CI. Classes sharing a super-
720
script do not differ significantly (p > 0.05).
29